21 February 2003 Internet service providers (ISPs) are stepping up action against spammers, with both America Online (AOL) and Microsoft pledging new spam blocking tools and even taking legal action against spammers.
AOL, the world’s biggest ISP with an estimated 35 million account holders, has set-up an anti-spam task force and promised to introduce a range of new anti-spam tools and lobby for tough new laws against spammers.
It already intercepts more than 750 million pieces of unsolicited commercial email every day, which costs the company millions of dollars every year just to deal with. The cost of dealing with this is ultimately paid for by AOL’s users.
AOL’s pledge follows legal action started last week by software giant Microsoft. It is suing a number of spammers, which it claims conducted a five month ‘dictionary attack’ against its Hotmail and MSN services.
A dictionary attack involves sending emails to common names and addresses at particular ISPs and companies, such as john.smith@hotmail.com. Because Hotmail has so many account holders, a high number will end up in peoples’ in-boxes.
However, such an approach can also have the same effect as a denial of service attack on an organisation’s email server if it lacks the capacity to handle the tens of millions of spam emails sent to it.
Organisations including ISPs can already deploy a range of defences to reduce spam.
The easiest and cheapest is to blacklist incoming email by the originating IP address found in the message header. There are a number of free anti-spam blacklisting services they can use for this purpose, such as the Spam Prevention Early Warning System (SPEWS) and the Spamhaus Block List (SBL).
A number of companies also offer anti-spam services.
Brightmail uses a ‘spam harvesting’ technique. As soon as it receives new spam at one of its ‘spam traps’ – email addresses published on the Internet, which are scooped up by spammers’ automated email address harvesting tools – a signature is instantly sent to customers and automatically added to their Brightmail filter.
BTOpenworld offers Brightmail’s service free on all its accounts.
MessageLabs offers outsourced anti-spam scanning based on artificial intelligence. It scans messages for phrases commonly used in spam, such as “You have opted in” and “This is NOT unsolicited email”, applies a score to each and if that score exceeds a certain level blacklists the email as spam.
However, this approach can lead to a high number of false positives.
ISPs can help themselves by taking more robust action against spammers on their own networks, including terminating both email access and ‘spamvertised’ web sites as soon as they are identified. To do this, many will have to toughen up their acceptable usage policies.
In addition, they ought to be scanning their own networks regularly for the presence of misconfigured mail servers, which are often abused to send spam anonymously. Finally, they need to start applying pressure on spam-friendly ISPs in other parts of the world, particularly in South Korea, China and South and Central America where spammers are still welcome and abuse complaints largely ignored.
